The proposed study addresses the need to develop effective techniques for interviewing young children who have been the victims of crime or witnesses to criminal acts. Because lengthy involvement in the legal and social service system may increase the trauma experienced by young children, several agencies and professional organizations have drafted recommendations to limit the number of times a child witness is interviewed. However, little is known about the impact these procedures will have on the quality of young children's testimony. Specifically, this project will document changes in the content and style of testimony across repeated questions both within and between interviews. Children in three age groups (4-, 6-, and 8- year-olds) and adults will witness an interaction between two adults that contains both unambiguous and ambiguous events. Each subject will then answer questions about those events either immediately after the events and one week later, or only after a one-week delay. During each interview, all questions will be re- peated three times. With respect to the content of their testimonies, subject's accuracy for the central events and details of the events will be recorded as a function of interview condition, as well as their tendency to respond to questions about which they have no information. The probability of response change will be computed within and between interviews for each question type. In addition, changes in response style that might affect judgments of credibility will be coded, such as speaking rate and number of nonverbal responses. In summary, the proposed research will delineate the effects of repeated interviews on the accuracy and presentation style of testimony from children and adults. This knowledge will aid policymakers in their attempts to balance the needs of the courts with those of child witnesses.